Parikh, Parth

Background Intracranial occlusions in embolic stroke of undetermined source are histopathologically similar to cardiac sources of embolism. Whether patients with embolic stroke of undetermined source and intracranial occlusion benefit from anticoagulation is unknown. Methods A multicenter retrospective cohort of adults with cryptogenic stroke was queried for patients with proximal or medium/distal vessel occlusion. The primary outcome was recurrent ischemic stroke, which was assessed in unadjusted and adjusted Cox proportional hazards models and repeated following 1:1 propensity score matching and biweight kernel density matching. Results Of the 2328 patients who were followed over a median of 1.31 years (interquartile range, 0.34–2.85), 999 (42.6%) had an intracranial occlusion. When compared with patients without an intracranial occlusion, those with an occlusion had fewer atherosclerotic vascular risk factors, more severe symptoms, and less severe cerebral microvascular disease. The rate of recurrent stroke was similar between patients with versus without an intracranial occlusion (6.8%/year [95% CI, 5.7–8.2] versus 7.0%/year [95% CI 6.0–8.1]; adjusted hazard ratio [HR], 1.09 [95% CI, 0.77–1.55]). There was no association between occlusion and recurrent stroke in the adjusted propensity score matching (HR, 1.01 [95% CI, 0.64–1.59]) or kernel density models (HR, 0.95 [95% CI, 0.62–1.45]). There was no interaction between occlusion and treatment with anticoagulation, sex, age, or high‐risk sources of embolism for the primary outcome in the unmatched and matched analyses. Conclusions Intracranial occlusion in patients with embolic stroke of undetermined source is not associated with a greater risk of recurrence when compared with patients without an intracranial occlusion. There was no difference in rate of recurrent stroke with anticoagulation when stratified by presence or absence of occlusion.

Introduction: Permanent perioperative vision loss is caused by ischemic optic neuropathy (ION) or central retinal artery occlusion (CRAO). Whereas diffusion restriction of the optic nerve (ON) on brain magnetic resonance imaging has been previously reported in perioperative posterior ION (PION), there are no reports of ON diffusion restriction in patients diagnosed with acute perioperative CRAO. We present a case of perioperative CRAO to highlight this neuroimaging finding for neuroradiologists and neurologists. Case Report: A 71-year-old male without vascular risk factors underwent maxillary bilateral antrostomy and septoplasty for chronic sinusitis. Twenty to thirty minutes upon awakening, he complained of painless left eye vision loss. Ophthalmoscopic examination showed retinal whitening, segmented arterioles, and hyperemic disc. Brain MR-diffusion weighted imaging/apparent diffusion coefficient revealed ON diffusion restriction in the proximal segment. Despite attempted reperfusion, left eye remained with no light perception at 6 months. Patients undergoing nonocular surgeries who develop perioperative vision loss related to PION may exhibit ON diffusion restriction but usually have normal ophthalmoscopic findings. CRAO shows retinal whitening, edema, segmentation of arterioles, and cherry red spot on ophthalmoscopy. A recent study reported that ON diffusion restriction in nonperioperative CRAO cases has a sensitivity and specificity of 55% and 70% to 100%. Here, PION was initially considered based on imaging. However, given the neuro-ophthalmic findings, a proximal embolus in the central retinal artery, obstructing its entrance into the proximal ON was deemed more likely. Conclusion: We highlight that proximal ON diffusion restriction on brain magnetic resonance imaging can be diagnostic of proximal thromboembolic CRAO. Future studies should evaluate the diagnostic utility and accuracy of MR-diffusion weighted imaging/apparent diffusion coefficient in perioperative visual loss.